Abstract
Purpose :
To explore reflectance quad-detection adaptive optics scanning light ophthalmoscopy (AOSLO) for photoreceptor imaging.
Methods :
A custom AOSLO was used to capture retinal images using five detectors, one with a 0.75 Airy disk diameter (ADD) on-axis and four detectors to split an annulus into quadrants. Inner diameter (ID) and outer diameter (OD) of the annulus were changed to search for the optimal values for photoreceptor imaging. Stacks of ~1° AOSLO image sequences (200 frames) spanning the depth of the retina were captured in a normal human using 790 nm light (54 µW at the pupil). Twenty-five frames of each sequence were registered and averaged to improve the signal-to-noise ratio. Images with photoreceptor mosaic in focus was subjectively determined, and nine new images were calculated from the averaged quadrant images. One of these calculated images (dark-field) is the sum of the quadrant images, and the other 8, that we termed split-detection (SD) images, were the difference divided by the sum of: the left/right and top/bottom quadrants (horizontal & vertical SD); diagonally opposite (radial SD); and adjacent quadrants (azimuthal SD).
Results :
The photoreceptor mosaic inner segment can be visualized in all SD images for all annuli tested (5-15 ADD). Subjective evaluation of the images suggests that cone mosaic contrast peaks around 7- 9 ADD. No preferential directions were observed between the horizontal and vertical, the two radial, and the four azimuthal SD images. Contrast of azimuthal SD images was lower when compared to other SD images.
Conclusions :
Simultaneous quad-detection of reflected light can be used to image the cone photoreceptor inner-segment mosaic with safe light levels. Our initial findings indicate that there are optimal annuli dimensions and quadrant image combinations, and these need to be determined by imaging a larger subject population.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.